Latin American Debt: I Don't Think We Are in Kansas Anymore
Total Page:16
File Type:pdf, Size:1020Kb
CARLOS F. DIAZ-ALEJANDRO ColumbiaUniversity Latin American Debt: I Don't Think We Are in Kansas Anymore BLAMING VICTIMS is an appealing evasion of responsibility, especially when the victims are far from virtuous. But when sins are as heteroge- neous as those of the LatinAmerican regimes of 1980,one wondershow well the exemplary mass punishmentfits the alleged individualcrime. Most Latin Americaneconomies, for a varietyof domestic and external reasons, in 1980-81faced the need for reformand adjustmentto the new internationaleconomic environment.However, the response was slow, and policy errors continued to be made. Yet the incompetence and torporof policymakersdo not fully explain the depth of the depression of the early 1980s in Latin America and the mediocre outlook for recovery. This paper will argue that what could have been a serious but manageablerecession has turned into a major development crisis un- precedentedsince the early 1930smainly because of the breakdownof internationalfinancial markets and an abruptchange in conditions and rules for internationallending. The nonlinearinteractions between this unusual and persistent external shock and risky or faulty domestic policies led to a crisis of severe depthand length, one thatneither shocks I am gratefulto Barry Adams, Andres Bianchi, EdmarBacha, WilliamCline, Jose Antonio Ocampo, Alan Stoga, Guillermo Calvo, Maurice Obstfeld, Michael Klein, VahidNowshirvani, Laurence M. Weiss, MiguelRodriguez, Jeff Frieden,Ernest Stern, FranciscoLopes, Ernesto Zedillo, Alberto Fracchia,Manuel Marfan, and Ted Truman for much help. Members of the Brookings Panel also provided many constructive suggestions. 335 336 Brookings Papers on Economic Activity, 2:1984 nor bad policy alone could have generated. Large capital outflows, in most cases encouragedby unconditionalcurrency convertibility,pro- vided a particularlyexplosive environmentfor the interactionof external shocks and imperfectpolicies. To make this central argumentplausible (proof seems impossible), I will review evidence from six Latin American countries: Argentina, Brazil, Colombia,Mexico-which have the region'slargest populations and gross domestic products-and medium-sizedChile and Venezuela. These six countries togetherrepresent at least 80 percent of any signifi- cant Latin Americaneconomic aggregate.The growthperformances of the six countriesvaried before 1981.Their external circumstances were different, with some of the group being oil importers and others oil exporters. Policy styles rangedfrom decidedly interventionistto mili- tantly laissez faire. Reliance on external borrowingvaried. Clear-eyed hindsightshows that the extent of policy errorswas also differentfrom country to country. Yet all six had serious economic difficultiesduring 1982-83and faced a weak recovery during1984. I will attemptto imaginehow the futurelooked to these countriesin 1980. I will analyze and date the crisis, examiningthe various external shocks, domestic adjustmentpolicies, and resultingeconomic perfor- mance of the different economies. In an econometric interlude I will scrutinize separately three crucial relations in these economies: the importfunction, the export function, and the determinantsof the real exchange rate. The paper's last section will focus on financial variables whose performancecharacterizes the crisis of the early 1980s.A discussion of the contrast between "national" external debt and private external assets will emphasize the asymmetricattitudes that exist toward those two sides of the LatinAmerican balance sheet. One overlooks important policy issues and redistributiveeffects when both private and public Latin American external debts are lumped together and when private (and sometimes even public) assets abroadare ignored. These consid- erations indicate that the debt crisis is not just a North-Southissue; for several Latin Americancountries it is also an issue of the distributionof domestic income and wealth. Many observers have marveled at the more-or-lesspunctual servicing of LatinAmerican debts duringthe early 1980s,a performancein sharpcontrast with the 1930s;however, "coun- tries" do not decide whetheror not to service debt-individual political actors do. The paper addresses the issue of debt servicing with a Carlos F. Diaz-Alejandro 337 discussionof these actors' perceptionsof the costs andbenefits of active or passive default.The papercloses with a review of the meagerrewards to adjustmentduring 1982-83, likely futurescenarios, and modest policy proposals. The development crisis in Latin America is forcing a salutaryreex- aminationof the role of the publicand private sectors in capitalformation and other economic activities; receiving particular attention is the efficiency and welfare consequences of the many functionsassumed by public sectors duringthe last half century. Nevertheless, the 1982-83 crisis was centered predominantlyon the balance of payments and,on the balance of internationalindebtedness. The paper will, therefore, focus on variablesimpinging strongly on foreignexchange flows. Background to Crisis: Could It Have Been Foreseen? This section firstsketches the heterogeneouspre-1981 performances, external conditions, and policies of the six countries under study and then identifies from their pre-crisis history the vulnerabilitiesand ex- cesses that may have led to the troubles of the 1980s. Several serious weaknesses are found, particularlyin Argentina and Chile, the two countries most in favor with the internationalfinancial community around1980-81, but I arguethat even in those countries,the information available at the time suggested no crisis of the magnitudewitnessed during1982-84. PRE-CRISIS PERFORMANCE A perspective on pre-crisisperformance comes fromexamining trend growth rates for 1960-83 as well as for two consecutive seven-year periods: the prosperous one from 1966 through 1973 and the troubled one from 1973 through 1980.1Four countries registeredimpressive or respectablerates of GDP growthfor the whole period:Brazil, Colombia, 1. Space limitationspreclude the extensive representationof these trends, obtained from the usual semilogarithmicregressions. They were computed from basic data at constant prices obtained from the United Nations Economic Commissionfor Latin America,which in turnrelies on nationalstatistics; and fromthe InternationalMonetary Fund, International Financial Statistics and Direction of Trade, various issues. Andres Bianchikindly provided the data availableat the U.N. EconomicCommission for Latin America. 338 Brookings Papers on Economic Activity, 2:1984 Mexico, and Venezuela. The growth rate slowed during the second septenniumbut remainedreasonably fast. (Mexico hada burstof growth averaging8 percent per year from 1978 through 1981.) Argentinaand Chile have grown slowly at least since the late 1920s.Argentina perked up during 1966-73 only to do miserablyin the following period; Chile did somewhat better in the 1970s, but its growth was quite unstable. It appearsthat very fast and very slow growth are associated with insta- bility, both when the six countries are comparedfor the whole period and when growth rates of a given country are contrastedbetween the two septennia. With the exceptions of Colombiaand Venezuela, there is little evidence that the growthof the 1970sled to significantimprove- ments in income distributionor social harmony;on the contrary, the societies were left with internaldivisions not conducive to nimble and resolute domestic responses to externalshocks. Using gross fixedinvestment and GDP data, one can computemarginal capital-outputratios to obtain a rough index of the productivityof the investment. With three-yearaverages and investmentlagged one year, the ratios are as follows: 1961-63 to 1971-73 to 1971-73 1979-81 Argentina 4.4 11.1 Brazil 2.9 3.3 Chile 3.8 5.0 Colombia 3.1 3.3 Mexico 2.5 3.1 Venezuela 4.2 7.2 Brazil, Colombia,and Mexico-the fastest growers-had the lowest marginalcapital-output ratios, andthe ratiosincreased only slightlyfrom one period to the next. Argentina,Chile, and Venezuela not only had lower investment productivity (which could be due either to supply inefficienciesor to poor managementof aggregatedemand) throughout the years under study but also experienced a sharp decline in that productivityafter 1971-73. For Venezuela, disaggregationinto oil and non-oil sectors might yield a better picture for the non-oil sector; non- oil output has grown more than OPEC-restrictedpetroleum production since the early 1970s. The correlation for 1960-83 between the growth of GDP and the growth of the purchasingpower of exports is high for our sample of Carlos F. Diaz-Alejandro 339 countries.2Contrary to widespreadassertions, the trend growth of the purchasing power of exports improved in four of the six countries between 1966-73and 1973-80.It is also noteworthythat Mexican exports measured in current dollars grew faster than those of Korea between 1972-74and 1979-81. Coffee and oil prices did well duringthe 1970s, so there is a sharp difference in the behavior of the terms of trade for Colombia, Mexico, and Venezuela on the one hand and, on the other, those for Argentina, Brazil, and Chile, which declined during 1973-80. In the same period, export volume grew rapidlyexcept for Colombiaand Venezuela. Also during 1973-80 Argentinaand Chile made impressive gains in export volume; of the six countries, only Mexico showed faster growthin both its terms of trade and exports volume during1973-80. Since at least the 1960s most Latin American